Taken from the top 12.1 mbsf of Hole 1087A, 193 samples (Samples 175-1087A-1H-1, 0-1 cm, through 2H-3, 90-91 cm) were examined for their planktonic foraminiferal content. Based on the preliminary age determinations from shipboard biostratigraphy and magnetostratigraphy (Shipboard Scientific Party, 1998), we chose a sampling resolution of 5 cm (Core 1H) to 10 cm (Core 2H) to reach a time resolution close to 3 k.y.
Planktonic foraminifers were obtained by disintegrating a 10-cm3 sample and washing it over a 125-µm sieve. The samples were dried and examined under the binocular microscope, and planktonic foraminifers were identified to species level from splits of the 125-µm fraction. This size fraction was chosen for the purpose of comparison with previously published census counts at locations close to Hole 1087A (Giraudeau, 1993; Little et al., 1997; Niebler and Gersonde, 1998), as well as to promote the environmental significance of typically small polar and subpolar species (e.g., N. pachyderma (s) and Globigerina quinqueloba) that are particularly abundant on the continental shelf of the southern Benguela region (Giraudeau, 1993).
An average of 250 specimens per sample were identified following the classical taxonomic concepts of Bé (1967), Parker (1962), and Hemleben et al. (1989). In addition, the intergrades between Neogloboquadrina dutertrei and N. pachyderma (s) (P-D intergrade of Kipp, 1976) were combined with N. pachyderma (d). Finally, G. menardii ssp. and Globorotalia tumida were combined into the single taxonomic category G. menardii. With the exception of this last species group, all raw counts (Table T1) were transformed into relative abundances. Because of the scarcity of G. menardii in species assemblages at Site 1087, but considering its importance as a potential tracer of tropical/subtropical waters of Indian Ocean origin (e.g., Berger and Wefer, 1996), we conducted a separate count of this species on the total unsplit >125-µm fraction. We subsequently expressed G. menardii abundance as both concentration (number of specimens/gram bulk sediment) and accumulation rate (AR) (number of specimens/cm/k.y.) (Table T2) following the formula:
where
Finally, we evaluated the coiling ratio of Globorotalia truncatulinoides (expressed as the ratio of dextral specimens vs. the sum of dextral + sinistral specimens) in conducting a separate count of this species in the total unsplit >125-µm fraction.
The stratigraphic
framework is based upon the benthic foraminiferal 
18O
record of Cibicidoides
wuellerstorfi, which was correlated with the SPECMAP standard record
(Imbrie et al., 1984). Details of the analytical procedure are given elsewhere
(Pierre et al., Chap.
12, this volume). Twenty stable isotope events were identified
between 0 and 13 mbsf (Table T3;
Fig. F2). The
final age model was achieved by linear interpolation between age-control points.
